ABSTRACT
Multitemporal fully polarimetric synthetic aperture radar (SAR) images have been successfully used for land use change detection. In urban expansion monitoring, we are interested not only in changes in defined areas but also those that change from one type to another. This letter presents a supervised urban land cover change detection method based on a series of the normalized difference ratio operators, which are generated by the polarimetric descriptors from SAR observables and polarimetric decomposition. The k-nearest neighbour classifier, superpixel segmentation method, and linear discriminant analysis technique are introduced to improve the accuracy and efficiency of the experiments. Compared with several previous methods, the proposed method avoids the repetitive classification processing of the used polarimetric SAR (PolSAR) images and the selection of the optimum polarimetric descriptors. Real fully PolSAR images are used for experimental analyses and validation of the proposed method. The classification accuracies for the change classes can approximately reach 80%, which demonstrates the effectiveness and usefulness of the proposed method.
Acknowledgements
The authors would like to thank the anonymous reviewers for their instructive comments, which helped improve this letter.
Disclosure statement
No potential conflict of interest was reported by the authors.